Rousseau, Marie-Claude ORCID: https://orcid.org/0000-0001-5215-8086; El-Zein, Mariam; Conus, Florence; Parent, Marie-Élise ORCID: https://orcid.org/0000-0002-4196-3773 et Li, Jun (2011). Quality assessment of the computerized Quebec BCG vaccination registry and linkage with administrative databases : a pilot study In: Abstracts of the 3rd North American Congress of Epidemiology, June 21-24, 2011 Montreal, Canada, June 21-24, 2011, Montréal, Québec.
Ce document n'est pas hébergé sur EspaceINRS.Résumé
The BCG (Bacillus Calmette-Gue´rin) Vaccination Registry for the Canadian province of Quebec comprises some 4 million vaccination records
from 1926-1993. Its content, available in paper format, has recently been
computerized using optical character recognition. In this pilot study, we
aimed to: 1) compare the computerized database with the paper format, and;
2) determine the proportion of successful linkages with demographic and
medical administrative databases. For the first aim, about 0.1% of the BCG
records were systematically selected from the paper format. For each record, discrepancies with the database for any of 13 variables including
names, birth date, gender, and characteristics of vaccination were documented. Exact agreement was observed for 99.6% of the 4,987 sampled
records; no more than one error per record was present. For the second aim,
a random sample of 3,500 subjects born in 1961-1974 and vaccinated from
1970-1974 was selected from the computerized BCG registry. Using personal identifiers (names, father’s given name, sex, and birth date), separate
linkages were conducted with the provincial medical insurance registration
file (deterministic) and birth registry (probabilistic). The proportion of successful linkages was 69.5% with the medical insurance file and 77% with
the birth registry, and varied greatly by birth year. In conclusion, the computerized data of the BCG registry was of excellent quality. Linkage of the
BCG registry to administrative databases, as a first step to create a retrospective cohort, was feasible. The linkage method, birth year, and missing
values in personal identifiers impacted on linkage success across year.
Type de document: | Document issu d'une conférence ou d'un atelier |
---|---|
Informations complémentaires: | Présentation orale 114 American Journal of Epidemiology (2011) 173(11):S29 |
Mots-clés libres: | - |
Centre: | Centre INRS-Institut Armand Frappier |
Date de dépôt: | 30 mars 2024 22:23 |
Dernière modification: | 30 mars 2024 22:23 |
URI: | https://espace.inrs.ca/id/eprint/14318 |
Gestion Actions (Identification requise)
Modifier la notice |